INDEX

Explanations

statements or actions that are intentional and planned

oai_token-act-pair · gpt-3.5-turbo

terms related to intentional and deliberate actions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Tycoon

-0.83

 Solitaire

-0.78

abies

-0.76

stars

-0.75

iao

-0.70

 Dogs

-0.68

 Locked

-0.67

orph

-0.67

zona

-0.67

 Veter

-0.66

POSITIVE LOGITS

 intentional

0.98

 effort

0.90

 violation

0.87

 omission

0.86

 wrongdoing

0.85

 deliberate

0.82

 contradiction

0.79

 undertaking

0.76

 deception

0.76

 attempt

0.76

Activations Density 0.025%

statements or actions that are intentional and planned

terms related to intentional and deliberate actions

No Comments

No Known Activations